Evaluating the merits of Survey and Observational Data in National Election Studies
This PhD project will build on the existing 2015 internet module of the British Election Study (iBES) collected by YouGov that merges survey and observational social media data. The primary goal is methodological in nature, and focuses on evaluating the validity and robustness of each data source as a measure of political attitudes and behaviour.
Survey research is a long-standing method for collecting data that is widely used by social scientists in their analyses of human behaviour and attitudes. Social media data, drawn from individuals’ news feeds and networks offers an alternative but more unstructured method of collecting large quantities of data to address the same questions. Both methods are subject to problems of representation and bias. Combining the two means that we now have the potential to understand people far better by analysing both what they do (on the internet) with what they say.
Are there better ways to study attitudes and political preferences than through surveys? Can we find out new insights about political behaviour by studying the websites they use, the tweets they send and the posts they write? Using this “observational data” may allow us to improve on the reliability of surveys by studying mismatches between reported and actual political preferences and activities online.
This PhD will match survey respondents’ online activities; the websites they visit, and their social media data alongside survey responses (including but not limited to the British Election Study) to investigate the link between online activity and answers to YouGov social surveys.
September 2017 to September 2021
Prior to starting a PhD, Oliver worked as a Data Analyst for Sheffield City Council as part of the Troubled Families Data Team. Before this he studied Politics at Sheffield Hallam University before moving to the University of Sheffield to complete an MSc in Politics with Research Methods.
Office: G.45 Humanities Bridgeford St